[Olsr-users] Routing-Problem | olsr.org - 0.6.0 | weight | etx_ff | mode ether

Markus Kittenberger (spam-protected)
Fri Mar 30 01:58:59 CEST 2012


i guess weight never changed the metric, it just gave the final decision
between two routes of equal cost/hopcount. (but not toally sure never
really used it)

nearest equivalent is using lqmult, but it affects olsrd routing globally,
not just local as the weight likely was doing it,.

Markus

On Thu, Mar 29, 2012 at 8:46 PM, Michael Rack <
(spam-protected)> wrote:

>  Yes it should change the metric for this route. But it does not.
>
> Here is a example how the routing looks like from txt_info-Plugin:
>
> VIEW IS FROM INTERNET-ROUTER as specified in my last mail in section
> Network-Design / Schema...
>
> The focus is on 172.16.12.160/27 it should be routed via "bond0"
>
> ROUTE              HOST          Metric     Interface
> 172.16.12.32/27    172.16.12.4   1  1.000   bond5   # ROUTER B.1
> 172.16.12.96/28    172.16.12.4   2  2.000   bond5   # ROUTER B.2
>     hidden node on metric 3
> 172.16.12.128/27   172.16.12.4   4  4.000   bond5   # ROUTER C
> 172.16.12.160/27   172.16.12.4   5  5.128   bond5   # ROUTER D
>
> 172.16.13.160/27   172.16.12.5   1  1.000   bond0   # ROUTER A.1
> 172.16.14.32/27    172.16.12.5   2  2.000   bond0   # ROUTER A.2
> 172.16.13.32/27    172.16.12.5   3  3.000   bond0   # ROUTER A.3
>
> This route should only been selected if "bond0" failed.
>
> This is the route as it should be:
>
> ROUTE              HOST          Metric     Interface
> 172.16.12.32/27    172.16.12.4   1  1.000   bond5   # ROUTER B.1
> 172.16.12.96/28    172.16.12.4   2  2.000   bond5   # ROUTER B.2
>
> 172.16.13.160/27   172.16.12.5   1  1.000   bond0   # ROUTER A.1
> 172.16.14.32/27    172.16.12.5   2  2.000   bond0   # ROUTER A.2
> 172.16.13.32/27    172.16.12.5   3  3.000   bond0   # ROUTER A.3
> 172.16.12.128/27   172.16.12.5   4  4.000   bond0   # ROUTER C
> 172.16.12.160/27   172.16.12.5   5  5.000   bond0   # ROUTER D
>
>
> So the weight parameter does not apply. Is this a bug in OLSR?
>
> The weight parameter is activated on ROUTER C.
> ROUTER C sees ROUTER A.3 and ROUTER B.2.
>
> ROUTER C gives ROUTER A.3 a weight of 5 and ROUTER C weight of 10. So ROUTER
> A.3 should be selected.
>
>
> Liebe Grüße aus Freilassing,
>
> Michael Rack
> RSM Freilassing
> --
> RSM Freilassing                 Tel.: +49 8654 607110
> Nocksteinstr. 13                Fax.: +49 8654 670438
> D-83395 Freilassing            www.rsm-freilassing.de
>
>
> Am 29.03.2012 17:15, schrieb Markus Kittenberger:
>
> i guess the weight does/did only with hopcount metric,..
>
>  Markus
>
>  p.s. true 0.5.5 is a previous version, but this was very very very long
> ago. (-;
>
> On Thu, Mar 29, 2012 at 3:56 PM, Michael Rack <
> (spam-protected)> wrote:
>
>> Dear olsr-users!
>>
>> I have a bad behavior and hope to get some help.
>>
>> I run only P2P Links (AP & STA) via wireless-lan and direct ethernet
>> connections. Every node runs currently olsr version 0.6.0.
>>
>> The Problem:
>> I have several upstreams and like to define which upstrem to be used via
>> the weight parameter. In previous versions of olsr this worked fine (as i
>> can remember v0.5.5). Now it dosn't.
>>
>> My Settings:
>>
>>> Willingness             3
>>> LinkQualityLevel        2
>>> LinkQualityAlgorithm    "etx_ff"
>>> UseHysteresis   no
>>> TcRedundancy    2
>>> MprCoverage     1
>>>
>>> InterfaceDefaults
>>> {
>>>        Ip4Broadcast   255.255.255.2555
>>>        HelloInterval   2.0
>>>        HelloValidityTime   60.0
>>>        TcInterval  5.0
>>>        TcValidityTime  120.0
>>>        MidInterval 5.0
>>>        MidValidityTime 120.0
>>>        HnaInterval 5.0
>>>        HnaValidityTime 120.0
>>> }
>>>
>>> Interface "wan"
>>> {
>>>        Mode "ether"
>>>        Weight 10
>>> }
>>>
>>> Interface "vlan2"
>>> {
>>>        Mode "ether"
>>>        Weight 2
>>> }
>>>
>>> Interface "vlan4"
>>> {
>>>        Mode "ether"
>>>        Weight 5
>>> }
>>>
>>
>> But currently the wan-interface is selected. While everything goes via
>> "wan" the routing change to "vlan4" but after a few minutes the routing
>> switch back to "wan" and i get a bad packet-loss for 4 seconds while the
>> routing is switching back to "wan" - but i don't know why because both
>> links are up and running. I don't know where the packets get loosed.
>>
>> Currently Network-Design:
>>
>>           INTERNET-ROUTER
>>>                 |
>>>      / ------- / \ ------- \
>>>   ROUTER A.1           ROUTER B.1
>>>      |                     |
>>>   ROUTER A.2           ROUTER B.2
>>>      |                     |
>>>   ROUTER A.3               |
>>>      |                     |
>>>      \ --------\ / ------- /
>>>        "vlan4"  |   "wan"
>>>              ROUTER C
>>>                 |
>>>              ROUTER D
>>>                 |
>>>               CLIENT
>>>
>>
>>
>> MTR is done via ROUTER D and the internet. The routing decision is made
>> on Router C that is multihomed (A.3 & B.2).
>>
>> So what have i to do, to make WEIGHT working?
>>
>> Liebe Grüße aus Freilassing,
>>
>> Michael Rack
>> RSM Freilassing
>> --
>> RSM Freilassing                 Tel.: +49 8654 607110
>> Nocksteinstr. 13                Fax.: +49 8654 670438
>> D-83395 Freilassing            www.rsm-freilassing.de
>>
>>
>> --
>> Olsr-users mailing list
>> (spam-protected)
>> https://lists.olsr.org/mailman/listinfo/olsr-users
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.olsr.org/pipermail/olsr-users/attachments/20120330/4131e18f/attachment.html>


More information about the Olsr-users mailing list